Toward a Simplified Fluid-Sensitive MRI Protocol in Small Joints of the Hand in Early Arthritis Patients: Reliability between mDixon and Regular FSE Fat Saturation MRI Sequences

Presentation Format: Oral presentation.

Purpose or Learning Objective: Magnetic resonance imaging (MRI) facilitates early recognition of rheumatoid arthritis by depicting inflammation. Contrast-enhanced T1-weighted and T2-weighted fat-suppressed sequences are sensitive and thus recommended, but they are hampered by invasiveness, costs, and long scan time. We therefore introduced a modified Dixon sequence (mDixon) that is more patient friendly, reduces cost, and shortens scan times by 83%. However, it is not known if this mDixon sequence is reliable in relation to regular-MRI sequences with and without contrast (T1- and T2-weighted, respectively). We determined the reliability between regular-MRI sequences with and without contrast (T1- and T2-weighted, respectively) and mDixon-MRI in early arthritis patients.

Methods or Background: A total of 29 patients with early arthritis underwent regular fat-suppressed MRI (T1- and T2-weighted) and mDixon sequences of metacarpophalangeals 2 to 5 and wrist joints. Two readers scored erosions, osteitis, synovitis, and tenosynovitis. Intraclass correlation coefficients (ICCs) between readers and comparing the two sequences were studied. Spearman correlations were determined.

Results or Findings: Performance between the two readers with the regular-MRI sequences was good to excellent (ICCs all ≥ 0.88). The between-reader ICC was also good to excellent for the mDixon-MRI (ICCs all ≥ 0.76). Next, ICCs between the two sequences were investigated to determine the reliability of mDixon. ICCs were good to excellent for a total Rheumatoid Arthritis Magnetic Resonance Imaging Score (RAMRIS) of 0.87 (95% confidence interval [CI], 0.74–0.94); for erosions, 0.88 (95% CI, 0.69–0.95); and total inflammation score, 0.84 (95% CI, 0.69–0.82). The individual MRI inflammation scores had ICCs for osteitis of 0.97 (95% CI, 0.93–0.98); for tenosynovitis, 0.78 (95% CI, 0.58–0.89); and for synovitis, 0.57 (95% CI, 0.26–0.77). In addition, scores were highly correlated for total RAMRIS, erosions, and total MRI inflammation score (ρ = 0.82, ρ = 0.81, and ρ = 0.80, respectively).

Conclusion: Regular-MRI sequences and mDixon-MRI perform equally well, suggesting the mDixon sequence is reliable to detect joint inflammation. Thus this is the first step toward an simplified and abridged MRI protocol in small hand joints in patients with early arthritis. The ultimate goal will be implementation of this mDixon-MRI sequence. Validation in larger studies is warranted.

No conflict of interest has been declared by the author(s).

Publication History

Article published online:
03 June 2021